Automated graded prognostic assessment for patients with hepatocellular carcinoma using machine learning.

BACKGROUND: Accurate mortality risk quantification is crucial for the management of hepatocellular carcinoma (HCC); however, most scoring systems are subjective. PURPOSE: To develop and independently validate a machine learning mortality risk quantification method for HCC patients using standard-of-care clinical data and liver radiomics on baseline magnetic resonance imaging (MRI). METHODS: This retrospective study included all patients with multiphasic contrast-enhanced MRI at the time of diagnosis treated at our institution. Patients were censored at their last date of follow-up, end-of-observation, or liver transplantation date. The data were randomly sampled into independent cohorts, with 85% for development and 15% for independent validation. An automated liver segmentation framework was adopted for radiomic feature extraction. A random survival forest combined clinical and radiomic variables to predict overall survival (OS), and performance was evaluated using Harrell's C-index. RESULTS: A total of 555 treatment-naïve HCC patients (mean age, 63.8 years ± 8.9 [standard deviation]; 118 females) with MRI at the time of diagnosis were included, of which 287 (51.7%) died after a median time of 14.40 (interquartile range, 22.23) months, and had median followed up of 32.47 (interquartile range, 61.5) months. The developed risk prediction framework required 1.11 min on average and yielded C-indices of 0.8503 and 0.8234 in the development and independent validation cohorts, respectively, outperforming conventional clinical staging systems. Predicted risk scores were significantly associated with OS (p < .00001 in both cohorts). CONCLUSIONS: Machine learning reliably, rapidly, and reproducibly predicts mortality risk in patients with hepatocellular carcinoma from data routinely acquired in clinical practice. CLINICAL RELEVANCE STATEMENT: Precision mortality risk prediction using routinely available standard-of-care clinical data and automated MRI radiomic features could enable personalized follow-up strategies, guide management decisions, and improve clinical workflow efficiency in tumor boards. KEY POINTS: • Machine learning enables hepatocellular carcinoma mortality risk prediction using standard-of-care clinical data and automated radiomic features from multiphasic contrast-enhanced MRI. • Automated mortality risk prediction achieved state-of-the-art performances for mortality risk quantification and outperformed conventional clinical staging systems. • Patients were stratified into low, intermediate, and high-risk groups with significantly different survival times, generalizable to an independent evaluation cohort.

Impact of 18F-FDG PET Intensity Normalization on Radiomic Features of Oropharyngeal Squamous Cell Carcinomas and Machine Learning-Generated Biomarkers.

We aimed to investigate the effects of 18F-FDG PET voxel intensity normalization on radiomic features of oropharyngeal squamous cell carcinoma (OPSCC) and machine learning-generated radiomic biomarkers. Methods: We extracted 1,037 18F-FDG PET radiomic features quantifying the shape, intensity, and texture of 430 OPSCC primary tumors. The reproducibility of individual features across 3 intensity-normalized images (body-weight SUV, reference tissue activity ratio to lentiform nucleus of brain and cerebellum) and the raw PET data was assessed using an intraclass correlation coefficient (ICC). We investigated the effects of intensity normalization on the features' utility in predicting the human papillomavirus (HPV) status of OPSCCs in univariate logistic regression, receiver-operating-characteristic analysis, and extreme-gradient-boosting (XGBoost) machine-learning classifiers. Results: Of 1,037 features, a high (ICC ≥ 0.90), medium (0.90 > ICC ≥ 0.75), and low (ICC < 0.75) degree of reproducibility across normalization methods was attained in 356 (34.3%), 608 (58.6%), and 73 (7%) features, respectively. In univariate analysis, features from the PET normalized to the lentiform nucleus had the strongest association with HPV status, with 865 of 1,037 (83.4%) significant features after multiple testing corrections and a median area under the receiver-operating-characteristic curve (AUC) of 0.65 (interquartile range, 0.62-0.68). Similar tendencies were observed in XGBoost models, with the lentiform nucleus-normalized model achieving the numerically highest average AUC of 0.72 (SD, 0.07) in the cross validation within the training cohort. The model generalized well to the validation cohorts, attaining an AUC of 0.73 (95% CI, 0.60-0.85) in independent validation and 0.76 (95% CI, 0.58-0.95) in external validation. The AUCs of the XGBoost models were not significantly different. Conclusion: Only one third of the features demonstrated a high degree of reproducibility across intensity-normalization techniques, making uniform normalization a prerequisite for interindividual comparability of radiomic markers. The choice of normalization technique may affect the radiomic features' predictive value with respect to HPV. Our results show trends that normalization to the lentiform nucleus may improve model performance, although more evidence is needed to draw a firm conclusion.

Automated MRI liver segmentation for anatomical segmentation, liver volumetry, and the extraction of radiomics.

OBJECTIVES: To develop and evaluate a deep convolutional neural network (DCNN) for automated liver segmentation, volumetry, and radiomic feature extraction on contrast-enhanced portal venous phase magnetic resonance imaging (MRI). MATERIALS AND METHODS: This retrospective study included hepatocellular carcinoma patients from an institutional database with portal venous MRI. After manual segmentation, the data was randomly split into independent training, validation, and internal testing sets. From a collaborating institution, de-identified scans were used for external testing. The public LiverHccSeg dataset was used for further external validation. A 3D DCNN was trained to automatically segment the liver. Segmentation accuracy was quantified by the Dice similarity coefficient (DSC) with respect to manual segmentation. A Mann-Whitney U test was used to compare the internal and external test sets. Agreement of volumetry and radiomic features was assessed using the intraclass correlation coefficient (ICC). RESULTS: In total, 470 patients met the inclusion criteria (63.9±8.2 years; 376 males) and 20 patients were used for external validation (41±12 years; 13 males). DSC segmentation accuracy of the DCNN was similarly high between the internal (0.97±0.01) and external (0.96±0.03) test sets (p=0.28) and demonstrated robust segmentation performance on public testing (0.93±0.03). Agreement of liver volumetry was satisfactory in the internal (ICC, 0.99), external (ICC, 0.97), and public (ICC, 0.85) test sets. Radiomic features demonstrated excellent agreement in the internal (mean ICC, 0.98±0.04), external (mean ICC, 0.94±0.10), and public (mean ICC, 0.91±0.09) datasets. CONCLUSION: Automated liver segmentation yields robust and generalizable segmentation performance on MRI data and can be used for volumetry and radiomic feature extraction. CLINICAL RELEVANCE STATEMENT: Liver volumetry, anatomic localization, and extraction of quantitative imaging biomarkers require accurate segmentation, but manual segmentation is time-consuming. A deep convolutional neural network demonstrates fast and accurate segmentation performance on T1-weighted portal venous MRI. KEY POINTS: • This deep convolutional neural network yields robust and generalizable liver segmentation performance on internal, external, and public testing data. • Automated liver volumetry demonstrated excellent agreement with manual volumetry. • Automated liver segmentations can be used for robust and reproducible radiomic feature extraction.

LiverHccSeg: A publicly available multiphasic MRI dataset with liver and HCC tumor segmentations and inter-rater agreement analysis.

Accurate segmentation of liver and tumor regions in medical imaging is crucial for the diagnosis, treatment, and monitoring of hepatocellular carcinoma (HCC) patients. However, manual segmentation is time-consuming and subject to inter- and intra-rater variability. Therefore, automated methods are necessary but require rigorous validation of high-quality segmentations based on a consensus of raters. To address the need for reliable and comprehensive data in this domain, we present LiverHccSeg, a dataset that provides liver and tumor segmentations on multiphasic contrast-enhanced magnetic resonance imaging from two board-approved abdominal radiologists, along with an analysis of inter-rater agreement. LiverHccSeg provides a curated resource for liver and HCC tumor segmentation tasks. The dataset includes a scientific reading and co-registered contrast-enhanced multiphasic magnetic resonance imaging (MRI) scans with corresponding manual segmentations by two board-approved abdominal radiologists and relevant metadata and offers researchers a comprehensive foundation for external validation, and benchmarking of liver and tumor segmentation algorithms. The dataset also provides an analysis of the agreement between the two sets of liver and tumor segmentations. Through the calculation of appropriate segmentation metrics, we provide insights into the consistency and variability in liver and tumor segmentations among the radiologists. A total of 17 cases were included for liver segmentation and 14 cases for HCC tumor segmentation. Liver segmentations demonstrates high segmentation agreement (mean Dice, 0.95 ± 0.01 [standard deviation]) and HCC tumor segmentations showed higher variation (mean Dice, 0.85 ± 0.16 [standard deviation]). The applications of LiverHccSeg can be manifold, ranging from testing machine learning algorithms on public external data to radiomic feature analyses. Leveraging the inter-rater agreement analysis within the dataset, researchers can investigate the impact of variability on segmentation performance and explore methods to enhance the accuracy and robustness of liver and tumor segmentation algorithms in HCC patients. By making this dataset publicly available, LiverHccSeg aims to foster collaborations, facilitate innovative solutions, and ultimately improve patient outcomes in the diagnosis and treatment of HCC.

MR Imaging-Based In Vivo Macrophage Imaging to Monitor Immune Response after Radiofrequency Ablation of the Liver.

PURPOSE: To establish molecular magnetic resonance (MR) imaging instruments for in vivo characterization of the immune response to hepatic radiofrequency (RF) ablation using cell-specific immunoprobes. MATERIALS AND METHODS: Seventy-two C57BL/6 wild-type mice underwent standardized hepatic RF ablation (70 °C for 5 minutes) to generate a coagulation area measuring 6-7 mm in diameter. CD68+ macrophage periablational infiltration was characterized with immunohistochemistry 24 hours, 72 hours, 7 days, and 14 days after ablation (n = 24). Twenty-one mice were subjected to a dose-escalation study with either 10, 15, 30, or 60 mg/kg of rhodamine-labeled superparamagnetic iron oxide nanoparticles (SPIONs) or 2.4, 1.2, or 0.6 mg/kg of gadolinium-160 (160Gd)-labeled CD68 antibody for assessment of the optimal in vivo dose of contrast agent. MR imaging experiments included 9 mice, each receiving 10-mg/kg SPIONs to visualize phagocytes using T2∗-weighted imaging in a horizontal-bore 9.4-T MR imaging scanner, 160Gd-CD68 for T1-weighted MR imaging of macrophages, or 0.1-mmol/kg intravenous gadoterate (control group). Radiological-pathological correlation included Prussian blue staining, rhodamine immunofluorescence, imaging mass cytometry, and immunohistochemistry. RESULTS: RF ablation-induced periablational infiltration (206.92 µm ± 12.2) of CD68+ macrophages peaked at 7 days after ablation (P < .01) compared with the untreated lobe. T2∗-weighted MR imaging with SPION contrast demonstrated curvilinear T2∗ signal in the transitional zone (TZ) (186 µm ± 16.9), corresponsing to Iron Prussian blue staining. T1-weighted MR imaging with 160Gd-CD68 antibody showed curvilinear signal in the TZ (164 µm ± 3.6) corresponding to imaging mass cytometry. CONCLUSIONS: Both SPION-enhanced T2∗-weighted and 160Gd-enhanced T1-weighted MR imaging allow for in vivo monitoring of macrophages after RF ablation, demonstrating the feasibility of this model to investigate local immune responses.

Machine Learning Models for Prediction of Posttreatment Recurrence in Early-Stage Hepatocellular Carcinoma Using Pretreatment Clinical and MRI Features: A Proof-of-Concept Study.

BACKGROUND. Posttreatment recurrence is an unpredictable complication after liver transplant for hepatocellular carcinoma (HCC) that is associated with poor survival. Biomarkers are needed to estimate recurrence risk before organ allocation. OBJECTIVE. This proof-of-concept study evaluated the use of machine learning (ML) to predict recurrence from pretreatment laboratory, clinical, and MRI data in patients with early-stage HCC initially eligible for liver transplant. METHODS. This retrospective study included 120 patients (88 men, 32 women; median age, 60.0 years) with early-stage HCC diagnosed who were initially eligible for liver transplant and underwent treatment by transplant, resection, or thermal ablation between June 2005 and March 2018. Patients underwent pretreatment MRI and posttreatment imaging surveillance. Imaging features were extracted from postcontrast phases of pretreatment MRI examinations using a pretrained convolutional neural network. Pretreatment clinical characteristics (including laboratory data) and extracted imaging features were integrated to develop three ML models (clinical model, imaging model, combined model) for predicting recurrence within six time frames ranging from 1 through 6 years after treatment. Kaplan-Meier analysis with time to recurrence as the endpoint was used to assess the clinical relevance of model predictions. RESULTS. Tumor recurred in 44 of 120 (36.7%) patients during follow-up. The three models predicted recurrence with AUCs across the six time frames of 0.60-0.78 (clinical model), 0.71-0.85 (imaging model), and 0.62-0.86 (combined model). The mean AUC was higher for the imaging model than the clinical model (0.76 vs 0.68, respectively; p = .03), but the mean AUC was not significantly different between the clinical and combined models or between the imaging and combined models (p > .05). Kaplan-Meier curves were significantly different between patients predicted to be at low risk and those predicted to be at high risk by all three models for the 2-, 3-, 4-, 5-, and 6-year time frames (p < .05). CONCLUSION. The findings suggest that ML-based models can predict recurrence before therapy allocation in patients with early-stage HCC initially eligible for liver transplant. Adding MRI data as model input improved predictive performance over clinical parameters alone. The combined model did not surpass the imaging model's performance. CLINICAL IMPACT. ML-based models applied to currently underutilized imaging features may help design more reliable criteria for organ allocation and liver transplant eligibility.

Impact of Chemoembolic Regimen on Immune Cell Recruitment and Immune Checkpoint Marker Expression following Transcatheter Arterial Chemoembolization in a VX2 Rabbit Liver Tumor Model.

PURPOSE: To characterize the effects of commonly used transcatheter arterial chemoembolization (TACE) regimens on the immune response and immune checkpoint marker expression using a VX2 rabbit liver tumor model. MATERIALS AND METHODS: Twenty-four VX2 liver tumor-bearing New Zealand white rabbits were assigned to 7 groups (n = 3 per group) undergoing locoregional therapy as follows: (a) bicarbonate infusion without embolization, (b) conventional TACE (cTACE) using a water-in-oil emulsion containing doxorubicin mixed 1:2 with Lipiodol, drug-eluting embolic-TACE with either (c) idarubicin-eluting Oncozene microspheres (40 µm) or (d) doxorubicin-eluting Lumi beads (40-90 µm). For each therapy arm (b-d), a tandem set of 3 animals with additional bicarbonate infusion before TACE was added, to evaluate the effect of pH modification on the immune response. Three untreated rabbits served as controls. Tissue was harvested 24 hours after treatment, followed by digital immunohistochemistry quantification (counts/µm2 ± SEM) of tumor-infiltrating cluster of differentiation 3+ T-lymphocytes, human leukocyte antigen DR type antigen-presenting cells (APCs), cytotoxic T-lymphocyte-associated protein-4 (CTLA-4), and programmed cell death protein-1 (PD-1)/PD-1 ligand (PD-L1) pathway axis expression. RESULTS: Lumi-bead TACE induced significantly more intratumoral T-cell and APC infiltration than cTACE and Oncozene-microsphere TACE. Additionally, tumors treated with Lumi-bead TACE expressed significantly higher intratumoral immune checkpoint markers compared with cTACE and Oncozene-microsphere TACE. Neoadjuvant bicarbonate demonstrated the most pronounced effect on cTACE and resulted in a significant increase in intratumoral cluster of differentiation 3+ T-cell infiltration compared with cTACE alone. CONCLUSIONS: This preclinical study revealed significant differences in evoked tumor immunogenicity depending on the choice of chemoembolic regimen for TACE.

Improved performance and consistency of deep learning 3D liver segmentation with heterogeneous cancer stages in magnetic resonance imaging.

PURPOSE: Accurate liver segmentation is key for volumetry assessment to guide treatment decisions. Moreover, it is an important pre-processing step for cancer detection algorithms. Liver segmentation can be especially challenging in patients with cancer-related tissue changes and shape deformation. The aim of this study was to assess the ability of state-of-the-art deep learning 3D liver segmentation algorithms to generalize across all different Barcelona Clinic Liver Cancer (BCLC) liver cancer stages. METHODS: This retrospective study, included patients from an institutional database that had arterial-phase T1-weighted magnetic resonance images with corresponding manual liver segmentations. The data was split into 70/15/15% for training/validation/testing each proportionally equal across BCLC stages. Two 3D convolutional neural networks were trained using identical U-net-derived architectures with equal sized training datasets: one spanning all BCLC stages ("All-Stage-Net": AS-Net), and one limited to early and intermediate BCLC stages ("Early-Intermediate-Stage-Net": EIS-Net). Segmentation accuracy was evaluated by the Dice Similarity Coefficient (DSC) on a dataset spanning all BCLC stages and a Wilcoxon signed-rank test was used for pairwise comparisons. RESULTS: 219 subjects met the inclusion criteria (170 males, 49 females, 62.8±9.1 years) from all BCLC stages. Both networks were trained using 129 subjects: AS-Net training comprised 19, 74, 18, 8, and 10 BCLC 0, A, B, C, and D patients, respectively; EIS-Net training comprised 21, 86, and 22 BCLC 0, A, and B patients, respectively. DSCs (mean±SD) were 0.954±0.018 and 0.946±0.032 for AS-Net and EIS-Net (p<0.001), respectively. The AS-Net 0.956±0.014 significantly outperformed the EIS-Net 0.941±0.038 on advanced BCLC stages (p<0.001) and yielded similarly good segmentation performance on early and intermediate stages (AS-Net: 0.952±0.021; EIS-Net: 0.949±0.027; p = 0.107). CONCLUSION: To ensure robust segmentation performance across cancer stages that is independent of liver shape deformation and tumor burden, it is critical to train deep learning models on heterogeneous imaging data spanning all BCLC stages.

Thermal ablation alone vs thermal ablation combined with transarterial chemoembolization for patients with small (<3 cm) hepatocellular carcinoma.

PURPOSE: Thermal ablation (TA) and transarterial chemoembolization (TACE) may be used alone or in combination (TACE+TA) for the treatment of hepatocellular carcinoma (HCC). The aim of our study was to compare the time to tumor progression (TTP) and overall survival (OS) for patients who received TA alone or TACE+TA for HCC tumors under 3 cm. MATERIALS AND METHODS: This HIPAA-compliant IRB-approved retrospective analysis included 85 therapy-naïve patients from 2010 to 2018 (63 males, 22 females, mean age 62.4 ± 8.5 years) who underwent either TA alone (n = 64) or TA in combination with drug-eluting beads (DEB)-TACE (n = 18) or Lipiodol-TACE (n = 3) for locoregional therapy of early stage HCC with maximum tumor diameter under 3 cm. Kaplan-Meier analysis was performed using the log-rank test to assess TTP and OS. RESULTS: All TA and TACE+TA treatments included were technically successful. TTP was 23.0 months in the TA group and 22.0 months in the TACE+TA group. There was no statistically significant difference in TTP (p = 0.64). Median OS was 69.7 months in the TA group and 64.6 months in the TACE+TA group. There was no statistically significant difference in OS (p = 0.14). The treatment cohorts had differences in AFP levels (p = 0.03) and BCLC stage (p = 0.047). Complication rates between patient groups were similar (p = 0.61). CONCLUSION: For patients with HCC under 3 cm, TA alone and TACE+TA have similar outcomes in terms of TTP and OS, suggesting that TACE+TA may not be needed for these tumors unless warranted by tumor location or other technical consideration.